ADVERTISEMENT
Hemodynamic Effects of Valve Asymmetry in Sapien 3 Transcatheter Aortic Valves
Abstract: Background. Eccentric valve deployment after transcatheter aortic valve replacement (TAVR) has been associated with abnormal leaflet shear stresses that may accelerate structural valve deterioration (SVD). This phenomenon has not been studied in patients receiving Sapien 3 prostheses (Edwards Lifesciences). Methods. A retrospective cohort analysis of 100 patients who received Sapien 3 valves between 2013 and 2015 at a single institution was performed. Axial fluoroscopic images from the co-planar view were used to assess TAVR asymmetry, which was defined as a ratio of left-to-right valve heights ≤0.9 or ≥1.1. Transthoracic echocardiograms (TTEs) were obtained at follow-up to analyze peak and mean aortic valve (AV) gradients, paravalvular leak (PVL), and aortic insufficiency (AI). Results. Overall, 26 mm and 29 mm valves had greater asymmetry (45.2% and 46.9%) compared to 23 mm valves (21.2%; P=.06). There was no relationship between pre-TAVR eccentricity and post-TAVR asymmetry, but greater annular calcification was associated with a higher incidence of TAVR asymmetry. Although asymmetry was associated with higher mean and peak AV gradients among 23 mm and 26 mm valves at both 1-year and 2-year follow-up exams, these results did not reach significance. There were no significant differences in PVL or AI severity between asymmetric and symmetric valves. Conclusions. Asymmetric deployment of Sapien 3 valves is common, particularly among 26 mm and 29 mm prostheses. Overall, we detected a small increase in gradients in smaller prostheses, which could reflect early subclinical SVD. Longer follow-up will be necessary to determine the extent to which eccentricity is associated with clinically significant SVD.
J INVASIVE CARDIOL 2018;30(4):138-143.
Key words: aortic stenosis, paravalvular leaks, structural valve deterioration
Transcatheter aortic valve replacement (TAVR) is an effective therapy in the management of severe aortic stenosis for patients at intermediate or high risk with surgery.1-3 However, as the target patient population expands to include lower-risk groups and younger patients, there is growing concern surrounding the degree to which structural valve deterioration (SVD) among TAVR valves may impact durability over time.4 One potential contributor to SVD is TAVR prosthetic eccentricity, which has been shown to lead to leaflet malcoaptation.5,6 In vitro studies have demonstrated that when TAVR valves are deployed eccentrically, there is an increase in leaflet deformity and peak leaflet stresses.7,8 Over time, these repeated stresses may lead to worsened valvular performance, including greater mean and peak aortic valve (AV) gradients and an increase in the incidence and severity of paravalvular leak (PVL) and aortic regurgitation (AI).9 In the present study, we utilized TAVR asymmetry as a surrogate measure for prosthetic eccentricity to assess its effects on hemodynamic valvular performance in Sapien 3 transcatheter aortic valves (Edwards Lifesciences).
Methods
Study design. This is a retrospective cohort study involving 100 patients who received Sapien 3 valves from 2013 to 2015 at a single institution. The purpose of this study was to evaluate the effect of TAVR asymmetry on hemodynamic valvular performance. The degree of TAVR asymmetry was measured using a plain film angiogram of the co-planar view after valve deployment. Symmetric and asymmetric valves were then analyzed with respect to their effects on hemodynamic valvular performance, including mean AV gradient, peak AV gradient, degree of central AI, and degree of PVL. Hemodynamic valvular performance was assessed using transthoracic echocardiograms (TTEs) at three times over the span of a 2-year follow-up period: (1) prior to discharge from hospital at time of TAVR; (2) at year 1; and (3) at year 2. For the purposes of quantifying valvular performance, regurgitation described as trivial received a score of 1, trace received a score of 2, mild received a score of 3, moderate received a score of 4, and severe received a score of 5.
Measurement of TAVR asymmetry. TAVR asymmetry was measured using freeze-frame captures of the final prosthesis deployment in the co-planar view at the time of procedure (Figure 1). The fluoroscopic angle was standardized for all patients and captured when the left coronary sinus, right coronary sinus, and non-coronary sinus were aligned with one another. Asymmetry measurements were obtained using a blinded observer by measuring the heights of each side of the prosthesis and then rounding to the nearest tenth of a millimeter. Asymmetry was calculated using the ratio of the left-to-right valve heights and defined by a ratio ≤0.90 or ≥1.1. All valve height ratios between 0.90 and 1.1 were considered symmetric for the purposes of this study.
Measurement of annular eccentricity index and calcium severity. Pre-TAVR annular eccentricity was obtained using preoperative computed tomography (CT) angiography of the chest. The eccentricity index is defined as 1 – (minimum diameter/maximum diameter), where the diameter is a measure of the cross-sectional distance across the aortic annulus. Measurements were obtained using a blinded third-party cardiologist 
and rounded to the nearest tenth millimeter. Post-TAVR prosthetic eccentricity was not assessed. The annular calcium severity score, reported as either none, trace, mild, moderate, or severe, was quantified on a scale of 1-5 in order to evaluate its impact on TAVR asymmetry (Table 1).
Population. Patients with severe aortic stenosis who were at high surgical risk were referred for TAVR with the Sapien 3 valve. The major inclusion criteria were mean AV gradient >40 mm Hg or jet velocity >4.0 m/sec, initial aortic valve area ≤0.8 cm2 or indexed effective orifice area <0.5 cm2/m2, New York Heart Association functional class II or higher, and Society of Thoracic Surgeons (STS) risk score ≥4 (or heart team consensus of intermediate or high risk). The major exclusion criteria for TAVR were prior myocardial infarction within the past 30 days, congenital unicuspid or bicuspid aortic valve, aortic regurgitation >3+, complex coronary artery disease, and severe left ventricular dysfunction with left ventricular ejection fraction <20%. Only patients with available data in subsequent serial echocardiograms were included in the study.
Statistical analysis. Statistical analysis was performed using Excel (Microsoft Corporation). Annular eccentricity, TAVR asymmetry, and hemodynamic parameters were analyzed by valve size (23 mm, 26 mm, and 29 mm) to account for intrinsic differences based on prosthetic outflow area. Comparisons between means of peak AV gradient, mean AV gradient, central AI severity, and PVL severity were performed using a one-tailed independent Student’s t-test. Comparisons between differences over time (from time of deployment to year 2) within symmetric or asymmetric valves types were performed using a paired analysis. Comparisons of differences in incidence of asymmetry across all three valve sizes were derived using Chi-square tests. A secondary analysis was performed after exclusion of subjects with greater than mild PVL or AI to account for the expected amplification of gradients in these patients. Differences were considered significant at P-value <.05.
Results
Annular calcium severity and TAVR asymmetry. The annular calcium severity score was lower for symmetric valves compared to asymmetric valves across all sizes and reached significance in the 23 mm valve group (3.87 vs 4.29; P=.03) (Table 1).
Annular eccentricity and TAVR asymmetry. Asymmetry was present in 36/100 patients (36.0%). The 26 mm and 29 mm valves exhibited a higher proportion of asymmetry compared with 23 mm valves (45.2% and 46.9% vs 21.2%, respectively; P=.06). The mean annular eccentricity indexes among asymmetric valves were 0.25, 0.19, and 0.22 for 23 mm, 26 mm and 29 mm valves, respectively. There were no significant differences in the annular eccentricity indices (EIs) between valves that exhibited asymmetry vs valves that did not.
Hemodynamic performance. Complete TTE data were available for 100%, 77%, and 43% of all patients before discharge, at 1-year follow-up, and at 2-year follow-up, respectively. The effects of asymmetry on hemodynamic performance, including mean AV gradient, peak AV gradient, PVL severity, and AI severity, are summarized in Tables 2 and 3. Overall, asymmetry had mixed effects on hemodynamic 
performance, particularly in the primary analysis, which included subjects with greater than mild AI or PVL. However, when these subjects were excluded, there was a general trend toward greater mean and peak AV gradients when asymmetry was present among the 23 mm and 26 mm valves. On the other hand, 29 mm valves generally tended to exhibit higher mean and peak AV pressures when symmetry was present, rather than asymmetry. Overall, in the paired analysis, which included only subjects with complete data from baseline to year 2, mean AV gradients tended to decrease over time in both asymmetric and symmetric valves. The exceptions to this were asymmetric 23 mm valves and symmetric 29 mm valves, which both experienced a non-significant increase in mean AV gradient from time of discharge to year 2. There were no statistically significant differences in mean PVL and AI between symmetric and asymmetric valves across the entire study period and all valve sizes.
Discussion
Previous studies of prosthetic eccentricity. In the present study, TAVR asymmetry, defined by a valve height ratio ≤0.90 or ≥1.0 as measured in the co-planar view, was used as a surrogate for prosthetic eccentricity to examine its effects on hemodynamic performance. Although TAVR asymmetry using plain film angiography has not been previously validated as a surrogate measure for prosthetic eccentricity, it is plausible that differences in axial heights and asymmetric expansion translate to changes in underlying orifice eccentricity.9 With respect to prosthetic eccentricity, there are several mechanisms by which greater eccentricity may lead to SVD. Computer-generated models and in vitro studies have demonstrated that prosthetic eccentricity is associated with increased leaflet stress and strain, which in turn may negatively impact valve hemodynamics (Table 3).10 Using a computational fluid dynamics model, Sun et al evaluated various degrees of prosthetic eccentricity on valve strain and leakage while under a typical transvalvular pressure of 120 mm Hg. They found that as eccentricity increased from 0.0 to 0.68, peak stress and strain increased by 58.58% and 10.95%, respectively. Similarly, Gunning et al used dual-camera stereophotogrammetry to further characterize strains patterns in the coaptation and commissure regions of TAVR leaflets. Compared with non-eccentric valves, eccentric valves (mean eccentricity index, 28%) were more likely to exhibit malcoaptation and adverse bending of the leaflet during peak diastole (13.76 ± 2.04% in eccentric vs 11.77 ± 1.61% in circular; P<.001).8,11 Furthermore, eccentric deployments were associated with a 2% increase in peak commissure strains, which correlated to an increase in peak stress from 390 kPa to 700 kPa. Abbasi et al demonstrated that in the presence of valve under-sizing, eccentric valves exhibited >100% increase in maximal principal stress at the commissure regions compared with non-eccentric valves.12 Lastly, although limited data exist, computational models have also demonstrated that eccentric valves are associated with adverse hemodynamic performance. Sun et al showed that when Sapien XT-like models were exposed to increasingly eccentric configurations in vitro, central AI significantly increased beyond an eccentricity >0.50.10 Young et al further quantified central AI using elliptical valve configurations in vitro and found that eccentric valves were associated with a significant larger fraction of regurgitant volume than circular valves (11.9 ± 3.3% vs 6.7 ± 1.3%; P<.001).13
In vivo studies have also demonstrated a relationship between eccentricity and declines in valvular performance (Table 4). When looking specifically at annular eccentricity, Wong et al showed that an eccentricity index >0.25 correlated with significant PVL (> grade II) (area under the curve [AUC], 0.834; P<.001) among 84 patients with CoreValves.14 Furthermore, significant PVL could be predicted with 80% sensitivity and 86% specificity if annular eccentricity exceeded 0.25. Expanding on these results, Di Martino evaluated whether TAVR eccentricity could predict PVL. They found that among 223 patients who received a CoreValve, stent frame eccentricity correlated well with incidence of significant PVL (P=.52; P<.001).15 In addition, using the Valve Academic Research Consortium (VARC) criteria, eccentric prosthetics demonstrated a greater median VARC-2 value compared to non-eccentric valves (12% vs 0%).
Greater central AI has also been linked to prosthetic eccentricity. For example, Delgado et al demonstrated that eccentric valves are more likely to exhibit moderate AI compared with circular valves (0.14 ± 0.02 vs 0.05 ± 0.03; P<.001).16 Rodriguez-Olivares et al also found that more eccentric CoreValves were associated with greater AI.9 Furthermore, Harrison et al showed that balloon postdilation following implantation of the CoreValve is associated with a reduction in post-TAVR AI without an increased incidence in ischemic cerebrovascular events, thus suggesting the importance of prosthesis circulatory in maintaining valvular performance.17 Although limited, these studies highlight the potential early association between prosthetic eccentricity and SVD.
Present study findings. In the present study, we found that asymmetry was more common among 26 mm and 29 mm valves compared to 23 mm valves (45.16% and 46.88% vs 21.21%; P=.06). Not surprisingly, greater annular calcification was associated with a significantly greater incidence of asymmetry, particularly in the 23 mm valve group. In fact, across all valve sizes, patients with prostheses that exhibited asymmetry tended to have greater annular calcification scores. These findings corroborate prior studies that have observed a relationship between greater annular calcification and an increased incidence of prosthetic eccentricity.15 Gooley et al, for example, demonstrated that among patients who received Lotus valves with EI >0.1, there was a significantly higher incidence of native annular calcification compared to prosthesis with EI <0.1 (0.24 vs 0.19; P=.01).18 Similarly, Bekeredjian et al demonstrated that the extent of native aortic annulus calcification is predictive of postprocedural prosthesis eccentricity in patients with CoreValves (r=0.48; P<.001) and that the extent of eccentricity in these patients is significantly associated with postprocedural PVL.19 This relationship between annular calcification and TAVR eccentricity is particularly clinically relevant as Wang et al has previously established that peak stresses are significantly greater at regions of native calcium deposits in TAVR valves.20 Thus, given our findings, it is feasible that more asymmetric valves are vulnerable to greater peak stresses, particularly at sites of preexisting annular calcification.
With respect to hemodynamic performance, we found no significant and consistent association between prosthetic asymmetry and valve performance after 2 years of follow-up. There were nonetheless subtle trends over the follow-up period that may suggest early signs of subclinical early structural valve deterioration and thus warrant the need for continued surveillance. For instance, 23 mm and 26 mm valves that exhibited asymmetry were associated with greater mean and peak AV gradients at both 1-year and 2-year follow-up compared with symmetric valves. Additionally, whereas symmetric 23 mm valves demonstrated a significant decrease in mean and peak AV gradients from time of placement to year 2, asymmetric 23 mm valves did not, and in fact exhibited a non-significant increase in gradients over time. Although premature, these changes in peak and mean pressure may suggest an early signal of SVD over longer-term follow-up, eg, beyond 5 or 10 years of use.
Study limitations. Several limitations of the present study exist. First, the follow-up period was limited to 2 years, and there was a high lost-to-follow-up rate by year 2 (47%). As such, it is possible that the postulated adverse effects of valve asymmetry may have been more apparent by 2 years had there been a greater population to examine. Second, it is unclear whether TAVR asymmetry predicts or even correlates with prosthetic eccentricity. In the present study, post-TAVR CT angiography was not available, and as such could not be assessed in relationship to prosthetic asymmetry. However, angiographic asymmetry is more easily obtained and does not require additional testing or radiation exposure. Last, there was a lack of standardization for predilation and postdilation, which may have implications for the incidence of valve eccentricity, although not for its hemodynamic effects.
Conclusions
In our study, TAVR asymmetry as obtained through plain film angiography was used as a surrogate for prosthetic eccentricity to predict valvular hemodynamic performance over a 2-year period. Overall, we found a non-significant trend in higher mean and peak AV gradients at 1-year and 2-year follow-up among 23 mm and 26 mm valves, but not 29 mm valves. No significant association was observed between TAVR asymmetry and central AI or PVL. There was a significant association between annular calcification severity and asymmetry in the 23 mm valve group, as well as higher non-significant degrees of annular calcification across all valve sizes that exhibited asymmetry. Future efforts should work to validate the use of plain film angiography as a surrogate marker for TAVR eccentricity via direct comparison to CT imaging. Long-term data will be necessary to fully expand on these preliminary results and to determine if valve asymmetry heralds prolonged deleterious effects on hemodynamic performance over time.
References
1. Leon MB, Smith CR, Mack M, et al. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med. 2010;363:1597-1607.
2. Leon MB, Smith CR, Mack MJ, et al. Transcatheter or surgical aortic-valve replacement in intermediate-risk patients. N Engl J Med. 2016;374:1609-1620.
3. Adams DH, Popma JJ, Reardon MJ, et al. Transcatheter aortic-valve replacement with a self-expanding prosthesis. N Engl J Med. 2014;370:1790-1798.
4. Webb JG, Dvir D. Is transcatheter aortic valve replacement a durable therapeutic strategy? JACC Cardiovasc Interv. 2015;8:1092-1094.
5. Mack MJ, Leon MB, Smith CR, et al. 5-year outcomes of transcatheter aortic valve replacement or surgical aortic valve replacement for high surgical risk patients with aortic stenosis (PARTNER 1): a randomised controlled trial. Lancet. 2015;385:2477-2484.
6. Gurvitch R, Wood DA, Tay EL, et al. Transcatheter aortic valve implantation: durability of clinical and hemodynamic outcomes beyond 3 years in a large patient cohort. Circulation. 2010;122:1319-1327.
7. Sun W, Li K, Sirois E. Simulated elliptical bioprosthetic valve deformation: implications for asymmetric transcatheter valve deployment. J Biomech. 2010;43:3085-3090.
8. Gunning PS, Saikrishnan N, Yoganathan AP, McNamara LM. Total ellipse of the heart valve: the impact of eccentric stent distortion on the regional dynamic deformation of pericardial tissue leaflets of a transcatheter aortic valve replacement. J R Soc Interface. 2015;12:20150737.
9. Rodriguez-Olivares R, Rahhab Z, Faquir NE, et al. Differences in frame geometry between balloon-expandable and self-expanding transcatheter heart valves and association with aortic regurgitation. Rev Esp Cardiol (Engl Ed). 2016;69:392-400.
10. Sun W, Li K, Sirois E. Simulated elliptical bioprosthetic valve deformation: Implications for asymmetric transcatheter valve deployment. J Biomech. 2010;43:3085-3090.
11. Gunning PS, Saikrishnan N, McNamara LM, Yoganathan AP. An in vitro evaluation of the impact of eccentric deployment on transcatheter aortic valve hemodynamics. Ann Biomed Eng. 2014;42:1195-1206.
12. Abbasi, M, Azadani, AN. TCT-621. The synergistic impact of eccentric and incomplete stent deployment on transcatheter aortic valve leaflet stress distribution. J Am Coll Cardiol. 2015;66:B253-B254.
13. Young E, Chen JF, Dong O, Gao S, Massiello A, Fukamachi K. Transcatheter heart valve with variable geometric configuration: in vitro evaluation. Artif Organs. 2011;35:1151-1159.
14. Wong DT, Bertaso AG, Liew GY, et al. Relationship of aortic annular eccentricity and paravalvular regurgitation post transcatheter aortic valve implantation with CoreValve. J Invasive Cardiol. 2013;25:190-195.
15. Di Martino LFM, Soliman OII, van Gils L, et al. Relation between calcium burden, echocardiographic stent frame eccentricity and paravalvular leakage after CoreValve transcatheter aortic valve implantation. Eur Heart J Cardiovasc Imaging. 2017;18:648-653.
16. Delgado V, Ng AC, van de Veire NR, et al. Transcatheter aortic valve implantation: role of multidetector row computed tomography to evaluate prosthesis positioning and deployment in relation to valve function. Eur Heart J. 2010;31:1114-1123.
17. Harrison JK, Hughes GC, Reardon MJ, et al. Balloon postdilation following implantation of a self-expanding transcatheter aortic valve bioprosthesis. JACC Cardiovasc Interv. 2017;10:168-175.
18. Gooley RP, Cameron JD, Meredith IT. Assessment of the geometric interaction between the Lotus transcatheter aortic valve prosthesis and the native ventricular aortic interface by 320-multidetector computed tomography. JACC Cardiovasc Interv. 2015;8:740-749.
19. Bekeredjian R, Bodingbauer D, Hofmann NP, et al. The extent of aortic annulus calcification is a predictor of postprocedural eccentricity and paravalvular regurgitation: a pre- and postinterventional cardiac computed tomography angiography study. J Invasive Cardiol. 2015;27:172-180.
20. Wang Q, Sirois E, Sun W. Patient-specific modeling of biomechanical interaction in transcatheter aortic valve deployment. J Biomech. 2012;45:1965-1971.
21. Martin C, Sun W. Comparison of transcatheter aortic valve and surgical bioprosthetic valve durability: a fatigue simulation study. J Biomech. 2015;48:3026-3034.
22. Caudron J, Fares J, Hauville C, et al. Evaluation of multislice computed tomography early after transcatheter aortic valve implantation with the Edwards Sapien bioprosthesis. Am J Cardiol. 2011;108:873-881.
23. Binder RK, Webb JG, Toggweiler S, et al. Impact of post-implant Sapien XT geometry and position on conduction disturbances, hemodynamic performance, and paravalvular regurgitation. JACC Cardiovasc Interv. 2013;6:462-468.
24. Di Martino LF, Vletter WB, Ren B, et al. Prediction of paravalvular leakage after transcatheter aortic valve implantation. Int J Cardiovasc Imaging. 2015;31:1461-1468.
25. Schuhbaeck A, Weingartner C, Arnold M, et al. Aortic annulus eccentricity before and after transcatheter aortic valve implantation: comparison of balloon-expandable and self-expanding prostheses. Eur J Radiol. 2015;84:1242-1248.
From the 1Division of Cardiovascular Medicine and 2Division of Cardiovascular Surgery, University of Pennsylvania, Philadelphia, Pennsylvania; and 3Howard University College of Medicine, School of Medicine, Washington, DC.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Herrmann reports consulting fees from Edwards Lifesciences; institutional research support from Edwards Lifesciences, Abbott Vascular, Medtronic, St. Jude Medical, and Boston Scientific. The remaining authors report no conflicts of interest regarding the content herein.
Manuscript submitted January 13, 2018, provisional acceptance given January 25, 2018, final version accepted February 12, 2018.
Address for correspondence: Howard C. Herrmann, MD, University of Pennsylvania, PCAM South Pavilion 11-107, 3400 Civic Center Blvd, Philadelphia, PA 19104. Email: Howard.herrmann@uphs.upenn.edu
